Repository navigation
Conversation
- A channel send uses GROUP_DATA. `im3:` text is rewritten to `im4:` before the row is saved. - Direct messages send v3 text. - The input bar can attach a drawing. The bubble draws a decoded image, and a bad payload shows the error text.
- Cover Send admits the drawing immediately, so the staged preview is gone - Failures keep the pixels and use a title for connection, encode, or size - The cover shows progress while sending, and Clear and the canvas stay idle - The Pixel Paint strings are translated in the other locales
- A channel MCOimg with no packet sender time matches a reaction by a hash of the datagram bytes, with no five-minute window. The row keeps the sender name from send time. - An incoming channel MCOimg joins the same receive path as channel text. It carries the decoded name and no wire time. - A channel send is classified once. The pool, the radio command, and the queue share that result instead of reading the text again. - A send that is not admitted leaves the draft or the pixels in place. A user resend changes the nonce before the queue stores the text.
A channel MCOimg has no sender timestamp, so extra paths join on the datagram data field. - Decrypt a group datagram as a data type and blob, and leave the sender timestamp empty. - Use the earliest matching RX row for the hop list, route, and region. Later rows add paths. Group text at wire time 0 stays a separate row. - On import, merge the same incoming picture for one radio and one channel. - Attach an RX row that was saved before the outgoing message.
- Show "Pixel drawing" instead of im4 text in notifications, the chat list, and VoiceOver - Hide the sent-time rows in Message Info for a channel image with no wire time - Send channel drawings as im3 group text on radios below firmware v11, which can't send group data - Match duplicate channel images only within five minutes, so a later identical drawing is no longer merged into an old row - Stamp send time on channel data so echoes of a delayed send still match - Bound channel and RX-log lookups by that window instead of scanning the whole channel - Treat text like "im2: on my way" as plain text unless the rest is Base91 - Draw raster images as one bitmap instead of one fill per pixel - Name palette colors and pixel colors for VoiceOver, and say that the channel size limit counts the node name - Restore the ChannelMessageEnvelope comment and fix the copy-matching rule in Reactions.md
- One ChannelReactionTarget on MessageDTO drives reaction indexing, lookup, sending, and admission for group text, outgoing rows, and datagram pictures - findChannelMessageForReaction falls back to the newest datagram copy in the channel, so the separate datagram query is gone - A datagram arrives as a .datagram kind framed as the group text it stands for, so sender Text Rules now match it like an im3 group-text picture - One datagram predicate replaces two; an im3 group-text picture sent at wire time 0 now takes the timed hash its sender uses - HeardRepeatsService shares one datagram observation step and takes the RX row in place of four context parameters - Incoming channel text with no sender no longer offers reactions it cannot send
- A picture's body is Base91 text, so a message rule could match random characters and block an unrelated picture - Sender rules still match a picture's sender in channels, DMs, and room posts - Add a test that covers both cases
- An incoming datagram MCOimg stores the receipt time and sets timestampCorrected, because the packet has no sender clock - The footer used that flag and showed the time as adjusted - Show the correction only when the message has a wire send time
Drawings could be lost to a stray tap, and a drawing that was too large was only reported after Send. - Show the encoded size against the limit under the canvas and disable Send when the drawing is too large - Add undo and redo, and make Clear undoable - Ask before Cancel discards a drawing that is not blank - Add eraser and fill tools, a grid toggle, and a preview - Replace the scroll view with a layout that fits the screen - Stop a stroke that leaves the canvas from joining the re-entry point - Make the canvas one VoiceOver element with a movable cursor
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds support for sending and decoding MCOimg v3 and v4. Sending MCOimg in channels is always done with v4 which uses GROUP_DATA. Direct messages use v3 which is plain text since DMs don't support GROUP_DATA.
Includes a basic pixel painter to draw images.
Closes #379